@aiready/pattern-detect 0.11.37 → 0.11.38

package/dist/index.js

@@ -341,7 +341,7 @@ function filterBySeverity(duplicates, minSeverity) {
   });
 }
 
-// src/detector.ts
+// src/core/extractor.ts
 function categorizePattern(code) {
   const lower = code.toLowerCase();
   if (lower.includes("request") && lower.includes("response") || lower.includes("router.") || lower.includes("app.get") || lower.includes("app.post") || lower.includes("express") || lower.includes("ctx.body")) {
@@ -386,17 +386,19 @@ function extractCodeBlocks(content, minLines) {
       currentBlock.push(line);
     }
     if (inFunction && braceDepth === 0 && currentBlock.length >= minLines) {
-      const blockContent = currentBlock.join("\n");
-      const linesOfCode = currentBlock.filter(
-        (l) => l.trim() && !l.trim().startsWith("//")
-      ).length;
-      blocks.push({
-        content: blockContent,
-        startLine: blockStart + 1,
-        endLine: i + 1,
-        patternType: categorizePattern(blockContent),
-        linesOfCode
-      });
+      const blockContent = currentBlock.join("\n").trim();
+      if (blockContent) {
+        const loc = currentBlock.filter(
+          (l) => l.trim() && !l.trim().startsWith("//")
+        ).length;
+        blocks.push({
+          content: blockContent,
+          startLine: blockStart + 1,
+          endLine: i + 1,
+          patternType: categorizePattern(blockContent),
+          linesOfCode: loc
+        });
+      }
       currentBlock = [];
       inFunction = false;
     } else if (inFunction && braceDepth === 0) {
@@ -406,15 +408,51 @@ function extractCodeBlocks(content, minLines) {
   }
   return blocks;
 }
+
+// src/core/normalizer.ts
 function normalizeCode(code) {
-  if (!code) {
-    return "";
-  }
+  if (!code) return "";
   return code.replace(/\/\/.*$/gm, "").replace(/\/\*[\s\S]*?\*\//g, "").replace(/"[^"]*"/g, '"STR"').replace(/'[^']*'/g, "'STR'").replace(/`[^`]*`/g, "`STR`").replace(/\b\d+\b/g, "NUM").replace(/\s+/g, " ").trim();
 }
+var stopwords = /* @__PURE__ */ new Set([
+  "return",
+  "const",
+  "let",
+  "var",
+  "function",
+  "class",
+  "new",
+  "if",
+  "else",
+  "for",
+  "while",
+  "async",
+  "await",
+  "try",
+  "catch",
+  "switch",
+  "case",
+  "default",
+  "import",
+  "export",
+  "from",
+  "true",
+  "false",
+  "null",
+  "undefined",
+  "this"
+]);
+function tokenize(norm) {
+  const punctuation = "(){}[];.,";
+  const cleaned = norm.split("").map((ch) => punctuation.includes(ch) ? " " : ch).join("");
+  return cleaned.split(/\s+/).filter((t) => t && t.length >= 3 && !stopwords.has(t.toLowerCase()));
+}
+
+// src/core/similarity.ts
 function jaccardSimilarity(tokens1, tokens2) {
   const set1 = new Set(tokens1);
   const set2 = new Set(tokens2);
+  if (set1.size === 0 && set2.size === 0) return 0;
   let intersection = 0;
   for (const token of set1) {
     if (set2.has(token)) intersection++;
@@ -422,6 +460,53 @@ function jaccardSimilarity(tokens1, tokens2) {
   const union = set1.size + set2.size - intersection;
   return union === 0 ? 0 : intersection / union;
 }
+
+// src/core/approx-engine.ts
+var ApproxEngine = class {
+  constructor(allBlocks, blockTokens) {
+    this.invertedIndex = /* @__PURE__ */ new Map();
+    this.allBlocks = allBlocks;
+    this.blockTokens = blockTokens;
+    this.buildIndex();
+  }
+  buildIndex() {
+    for (let i = 0; i < this.blockTokens.length; i++) {
+      for (const tok of this.blockTokens[i]) {
+        let arr = this.invertedIndex.get(tok);
+        if (!arr) {
+          arr = [];
+          this.invertedIndex.set(tok, arr);
+        }
+        arr.push(i);
+      }
+    }
+  }
+  findCandidates(blockIdx, minSharedTokens, maxCandidates) {
+    const block1 = this.allBlocks[blockIdx];
+    const block1Tokens = this.blockTokens[blockIdx];
+    const counts = /* @__PURE__ */ new Map();
+    const rareTokens = block1Tokens.filter((tok) => {
+      const freq = this.invertedIndex.get(tok)?.length || 0;
+      return freq < this.allBlocks.length * 0.1;
+    });
+    for (const tok of rareTokens) {
+      const ids = this.invertedIndex.get(tok);
+      if (!ids) continue;
+      for (const j of ids) {
+        if (j <= blockIdx) continue;
+        if (this.allBlocks[j].file === block1.file) continue;
+        counts.set(j, (counts.get(j) || 0) + 1);
+      }
+    }
+    return Array.from(counts.entries()).filter(([j, shared]) => {
+      const block2Size = this.blockTokens[j].length;
+      const minSize = Math.min(block1Tokens.length, block2Size);
+      return shared >= minSharedTokens && shared / minSize >= 0.3;
+    }).sort((a, b) => b[1] - a[1]).slice(0, maxCandidates).map(([j, shared]) => ({ j, shared }));
+  }
+};
+
+// src/detector.ts
 async function detectDuplicatePatterns(files, options) {
   const {
     minSimilarity,
@@ -435,274 +520,92 @@ async function detectDuplicatePatterns(files, options) {
   const duplicates = [];
   const maxComparisons = approx ? Infinity : 5e5;
   const allBlocks = files.flatMap(
-    (file) => extractCodeBlocks(file.content, minLines).filter((block) => block.content && block.content.trim().length > 0).map((block) => ({
-      content: block.content,
-      startLine: block.startLine,
-      endLine: block.endLine,
+    (file) => extractCodeBlocks(file.content, minLines).filter(
+      (block) => block && block.content && block.content.trim().length > 0
+    ).map((block) => ({
+      ...block,
       file: file.file,
       normalized: normalizeCode(block.content),
-      patternType: block.patternType,
-      tokenCost: (0, import_core2.estimateTokens)(block.content),
-      linesOfCode: block.linesOfCode
+      tokenCost: block.content ? (0, import_core2.estimateTokens)(block.content) : 0
     }))
   );
-  if (!options.onProgress) {
-    console.log(`Extracted ${allBlocks.length} code blocks for analysis`);
-  }
-  const pythonFiles = files.filter((f) => f.file.toLowerCase().endsWith(".py"));
+  const pythonFiles = files.filter((f) => f.file.endsWith(".py"));
   if (pythonFiles.length > 0) {
     const { extractPythonPatterns: extractPythonPatterns2 } = await Promise.resolve().then(() => (init_python_extractor(), python_extractor_exports));
-    const patterns = await extractPythonPatterns2(
+    const pythonPatterns = await extractPythonPatterns2(
       pythonFiles.map((f) => f.file)
     );
-    const pythonBlocks = patterns.filter((p) => p.code && p.code.trim().length > 0).map((p) => ({
-      content: p.code,
-      startLine: p.startLine,
-      endLine: p.endLine,
-      file: p.file,
-      normalized: normalizeCode(p.code),
-      patternType: p.type,
-      tokenCost: (0, import_core2.estimateTokens)(p.code),
-      linesOfCode: p.endLine - p.startLine + 1
-    }));
-    allBlocks.push(...pythonBlocks);
-    if (!options.onProgress) {
-      console.log(`Added ${pythonBlocks.length} Python patterns`);
-    }
-  }
-  if (!approx && allBlocks.length > 500) {
-    console.log(
-      `\u26A0\uFE0F  Using --no-approx mode with ${allBlocks.length} blocks may be slow (O(B\xB2) complexity).`
-    );
-    console.log(
-      `   Consider using approximate mode (default) for better performance.`
+    allBlocks.push(
+      ...pythonPatterns.map((p) => ({
+        content: p.code,
+        startLine: p.startLine,
+        endLine: p.endLine,
+        file: p.file,
+        normalized: normalizeCode(p.code),
+        patternType: p.type,
+        tokenCost: p.code ? (0, import_core2.estimateTokens)(p.code) : 0,
+        linesOfCode: p.endLine - p.startLine + 1
+      }))
     );
   }
-  const stopwords = /* @__PURE__ */ new Set([
-    "return",
-    "const",
-    "let",
-    "var",
-    "function",
-    "class",
-    "new",
-    "if",
-    "else",
-    "for",
-    "while",
-    "async",
-    "await",
-    "try",
-    "catch",
-    "switch",
-    "case",
-    "default",
-    "import",
-    "export",
-    "from",
-    "true",
-    "false",
-    "null",
-    "undefined",
-    "this"
-  ]);
-  const tokenize = (norm) => {
-    const punctuation = "(){}[];.,";
-    const cleaned = norm.split("").map((ch) => punctuation.includes(ch) ? " " : ch).join("");
-    return cleaned.split(/\s+/).filter((t) => t && t.length >= 3 && !stopwords.has(t.toLowerCase()));
-  };
   const blockTokens = allBlocks.map((b) => tokenize(b.normalized));
-  const invertedIndex = /* @__PURE__ */ new Map();
-  if (approx) {
-    for (let i = 0; i < blockTokens.length; i++) {
-      for (const tok of blockTokens[i]) {
-        let arr = invertedIndex.get(tok);
-        if (!arr) {
-          arr = [];
-          invertedIndex.set(tok, arr);
-        }
-        arr.push(i);
-      }
-    }
-  }
-  const totalComparisons = approx ? void 0 : allBlocks.length * (allBlocks.length - 1) / 2;
-  if (totalComparisons !== void 0) {
-    console.log(
-      `Processing ${totalComparisons.toLocaleString()} comparisons in batches...`
-    );
-  } else {
-    console.log(
-      `Using approximate candidate selection to reduce comparisons...`
-    );
-  }
+  const engine = approx ? new ApproxEngine(allBlocks, blockTokens) : null;
   let comparisonsProcessed = 0;
-  let comparisonsBudgetExhausted = false;
   const startTime = Date.now();
   for (let i = 0; i < allBlocks.length; i++) {
-    if (maxComparisons && comparisonsProcessed >= maxComparisons) {
-      comparisonsBudgetExhausted = true;
-      break;
-    }
+    if (maxComparisons && comparisonsProcessed >= maxComparisons) break;
     if (i % batchSize === 0 && i > 0) {
       if (options.onProgress) {
-        options.onProgress(i, allBlocks.length, `pattern-detect: analyzing blocks`);
+        options.onProgress(i, allBlocks.length, "Analyzing patterns");
       } else {
-        const elapsed = ((Date.now() - startTime) / 1e3).toFixed(1);
-        const duplicatesFound = duplicates.length;
-        if (totalComparisons !== void 0) {
-          const progress = (comparisonsProcessed / totalComparisons * 100).toFixed(1);
-          const remaining = totalComparisons - comparisonsProcessed;
-          const rate = comparisonsProcessed / parseFloat(elapsed);
-          const eta = remaining > 0 ? (remaining / rate).toFixed(0) : 0;
-          console.log(
-            `   ${progress}% (${comparisonsProcessed.toLocaleString()}/${totalComparisons.toLocaleString()} comparisons, ${elapsed}s elapsed, ~${eta}s remaining, ${duplicatesFound} duplicates)`
-          );
-        } else {
-          console.log(
-            `   Processed ${i.toLocaleString()}/${allBlocks.length} blocks (${elapsed}s elapsed, ${duplicatesFound} duplicates)`
-          );
-        }
+        const elapsed = (Date.now() - startTime) / 1e3;
+        console.log(
+          `   Processed ${i}/${allBlocks.length} blocks (${elapsed.toFixed(1)}s, ${duplicates.length} duplicates)`
+        );
       }
-      await new Promise((resolve) => setImmediate(resolve));
+      await new Promise((r) => setImmediate((resolve) => r(resolve)));
     }
     const block1 = allBlocks[i];
-    let candidates = null;
-    if (approx) {
-      const counts = /* @__PURE__ */ new Map();
-      const block1Tokens = new Set(blockTokens[i]);
-      const block1Size = block1Tokens.size;
-      const rareTokens = blockTokens[i].filter((tok) => {
-        const blocksWithToken = invertedIndex.get(tok)?.length || 0;
-        return blocksWithToken < allBlocks.length * 0.1;
-      });
-      for (const tok of rareTokens) {
-        const ids = invertedIndex.get(tok);
-        if (!ids) continue;
-        for (const j of ids) {
-          if (j <= i) continue;
-          if (allBlocks[j].file === block1.file) continue;
-          counts.set(j, (counts.get(j) || 0) + 1);
-        }
-      }
-      candidates = Array.from(counts.entries()).filter(([j, shared]) => {
-        const block2Tokens = blockTokens[j];
-        const block2Size = block2Tokens.length;
-        const minSize = Math.min(block1Size, block2Size);
-        const sharedPercentage = shared / minSize;
-        return shared >= minSharedTokens && sharedPercentage >= 0.3;
-      }).sort((a, b) => b[1] - a[1]).slice(0, Math.min(maxCandidatesPerBlock, 5)).map(([j, shared]) => ({ j, shared }));
-    }
-    if (approx && candidates) {
-      for (const { j } of candidates) {
-        if (!approx && maxComparisons !== Infinity && comparisonsProcessed >= maxComparisons) {
-          console.log(
-            `\u26A0\uFE0F  Comparison safety limit reached (${maxComparisons.toLocaleString()} comparisons in --no-approx mode).`
-          );
+    const candidates = engine ? engine.findCandidates(i, minSharedTokens, maxCandidatesPerBlock) : allBlocks.slice(i + 1).map((_, idx) => ({ j: i + 1 + idx, shared: 0 }));
+    for (const { j } of candidates) {
+      if (!approx && comparisonsProcessed >= maxComparisons) break;
+      comparisonsProcessed++;
+      const block2 = allBlocks[j];
+      if (block1.file === block2.file) continue;
+      const sim = jaccardSimilarity(blockTokens[i], blockTokens[j]);
+      if (sim >= minSimilarity) {
+        const severity = calculateSeverity(
+          block1.file,
+          block2.file,
+          block1.content,
+          sim,
+          block1.linesOfCode
+        );
+        const dup = {
+          file1: block1.file,
+          file2: block2.file,
+          line1: block1.startLine,
+          line2: block2.startLine,
+          endLine1: block1.endLine,
+          endLine2: block2.endLine,
+          similarity: sim,
+          snippet: block1.content.substring(0, 200),
+          patternType: block1.patternType,
+          tokenCost: block1.tokenCost,
+          linesOfCode: block1.linesOfCode,
+          severity: severity.severity,
+          reason: severity.reason,
+          suggestion: severity.suggestion
+        };
+        duplicates.push(dup);
+        if (streamResults)
           console.log(
-            `   This prevents excessive runtime on large repos. Consider using approximate mode (default) or --min-lines to reduce blocks.`
-          );
-          break;
-        }
-        comparisonsProcessed++;
-        const block2 = allBlocks[j];
-        const similarity = jaccardSimilarity(blockTokens[i], blockTokens[j]);
-        if (similarity >= minSimilarity) {
-          const { severity, reason, suggestion, matchedRule } = calculateSeverity(
-            block1.file,
-            block2.file,
-            block1.content,
-            similarity,
-            block1.linesOfCode
-          );
-          const duplicate = {
-            file1: block1.file,
-            file2: block2.file,
-            line1: block1.startLine,
-            line2: block2.startLine,
-            endLine1: block1.endLine,
-            endLine2: block2.endLine,
-            similarity,
-            snippet: block1.content.split("\n").slice(0, 5).join("\n") + "\n...",
-            patternType: block1.patternType,
-            tokenCost: block1.tokenCost + block2.tokenCost,
-            linesOfCode: block1.linesOfCode,
-            severity,
-            reason,
-            suggestion,
-            matchedRule
-          };
-          duplicates.push(duplicate);
-          if (streamResults) {
-            console.log(
-              `
-   \u2705 Found: ${duplicate.patternType} ${Math.round(similarity * 100)}% similar`
-            );
-            console.log(
-              `      ${duplicate.file1}:${duplicate.line1}-${duplicate.endLine1} \u21D4 ${duplicate.file2}:${duplicate.line2}-${duplicate.endLine2}`
-            );
-            console.log(
-              `      Token cost: ${duplicate.tokenCost.toLocaleString()}`
-            );
-          }
-        }
-      }
-    } else {
-      for (let j = i + 1; j < allBlocks.length; j++) {
-        if (maxComparisons && comparisonsProcessed >= maxComparisons) break;
-        comparisonsProcessed++;
-        const block2 = allBlocks[j];
-        if (block1.file === block2.file) continue;
-        const similarity = jaccardSimilarity(blockTokens[i], blockTokens[j]);
-        if (similarity >= minSimilarity) {
-          const { severity, reason, suggestion, matchedRule } = calculateSeverity(
-            block1.file,
-            block2.file,
-            block1.content,
-            similarity,
-            block1.linesOfCode
+            `[DUPLICATE] ${dup.file1}:${dup.line1} <-> ${dup.file2}:${dup.line2} (${Math.round(sim * 100)}%)`
           );
-          const duplicate = {
-            file1: block1.file,
-            file2: block2.file,
-            line1: block1.startLine,
-            line2: block2.startLine,
-            endLine1: block1.endLine,
-            endLine2: block2.endLine,
-            similarity,
-            snippet: block1.content.split("\n").slice(0, 5).join("\n") + "\n...",
-            patternType: block1.patternType,
-            tokenCost: block1.tokenCost + block2.tokenCost,
-            linesOfCode: block1.linesOfCode,
-            severity,
-            reason,
-            suggestion,
-            matchedRule
-          };
-          duplicates.push(duplicate);
-          if (streamResults) {
-            console.log(
-              `
-   \u2705 Found: ${duplicate.patternType} ${Math.round(similarity * 100)}% similar`
-            );
-            console.log(
-              `      ${duplicate.file1}:${duplicate.line1}-${duplicate.endLine1} \u21D4 ${duplicate.file2}:${duplicate.line2}-${duplicate.endLine2}`
-            );
-            console.log(
-              `      Token cost: ${duplicate.tokenCost.toLocaleString()}`
-            );
-          }
-        }
       }
     }
   }
-  if (comparisonsBudgetExhausted) {
-    console.log(
-      `\u26A0\uFE0F  Comparison budget exhausted (${maxComparisons.toLocaleString()} comparisons). Use --max-comparisons to increase.`
-    );
-  }
-  return duplicates.sort(
-    (a, b) => b.similarity - a.similarity || b.tokenCost - a.tokenCost
-  );
+  return duplicates;
 }
 
 // src/grouping.ts

package/dist/index.mjs

@@ -7,7 +7,7 @@ import {
   generateSummary,
   getSeverityLabel,
   getSmartDefaults
-} from "./chunk-YSDOUNJJ.mjs";
+} from "./chunk-6OEHUI5J.mjs";
 export {
   analyzePatterns,
   calculatePatternScore,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@aiready/pattern-detect",
-  "version": "0.11.37",
+  "version": "0.11.38",
   "description": "Semantic duplicate pattern detection for AI-generated code - finds similar implementations that waste AI context tokens",
   "main": "./dist/index.js",
   "module": "./dist/index.mjs",
@@ -45,7 +45,7 @@
   "dependencies": {
     "commander": "^14.0.0",
     "chalk": "^5.3.0",
-    "@aiready/core": "0.9.38"
+    "@aiready/core": "0.9.39"
   },
   "devDependencies": {
     "tsup": "^8.3.5",